A good UX design gains popularity because of its visual appeal and how easily users can find what they’re looking for. A website’s top usability problem is that content is structured based on what makes sense to the company and not the end-users.

Card sorting is one of the primary ways to sort out the navigation scheme or structure of an app, website, or prototype that best matches the users’ mental model or conditioning.

It represents how they navigate and expect content to be categorized and grouped.

This helps with quicker transactions and reduces the bounce rate on a website.

What is card sorting?

Card sorting is a UX research method that involves studying participants grouping individual labels written on notecards based on criteria that seem sensible to them.

It helps understand their attitudes, values, preferences, and behaviors related to the domain under study. This method helps uncover how the target audience’s domain knowledge is structured, and it serves to create an information architecture based on the users’ expectations.

This image has an empty alt attribute; its file name is Card-sorting_20201009_1-1024x731.png

Information Architecture defines how all the content of a site is related and fits together as a coherent unit. It refers to the structure of information on a website and is depicted with site maps, spreadsheets, and diagrams. Card Sorting helps understand the user’s perception of the information space in the IA planning stage.

The grouping or naming of objects and concepts may be represented on physical or virtual cards on computer screens; or photos in either physical or digital formats.

Sorting is usually performed by the potential users of an interactive solution providing,

  • Terminology (what people call things)
  • Relationships (proximity, similarity)
  • Categories (groups and their names)

This information is used by designers to decide which items should be grouped in displays, how the menu elements should be organized and labeled, and, more fundamentally, the words to describe the objects that need user attention.

Card Sorting is useful to:

  • Design a new website or enhance an existing website or section update
  • Discover how your customers expect to see your information grouped on your website.
  • Get the user’s familiarized with the information architecture and the user-centric design.
  • Prevent the confused organization of content on a website or webpage.

Benefits of Card Sorting for UX Designers

  • It is reliable, cost-efficient, and easy to set up.
  • Help in designing an optimized navigation structure for a website/app.
  • Promote smoother interactions between users and the system by presenting a user-friendly interface.
  • Help define a clear content strategy for a website.
  • Help scientifically evaluate and fix website structural problems in the early stages of design evolution.

When to use it?

Card sorting is used in the following situations. It is typically useful in improving the labeling, grouping, and organization of information.

  • Information or content has to be grouped by type, subject, or category.
  • Group your content or products in a way that will make sense to your user base.
  • Untangle complexity in information depiction (sitemaps).
  • Sort or organize content in a particular order (FAQ section).
  • Decide on a structure for your website.
  • Find the right words for navigation.

Types of Card Sorting Models

Open Card Sorting

In this method, participants receive specially prepared cards, sort them into the categories provided, create their self-defined categories, and name them. They are free to name groups they’ve created with the cards in the deck using labels that best describe them.

The open card is commonly used for new/existing information architectures, organizing products on a new site or app, or creating a new IA from scratch.

This image has an empty alt attribute; its file name is Card-sorting_20201009_2-1024x677.png

It helps explore how your target audience understands content headings, topics, and their tags. It is worth noting that users suggest each category based on how they think of these topics. Open sorting has the following advantages.

  • Promote participant-driven discoveries, limiting any preconceived barriers by the designer.
  • Lend the participants more flexibility.
  • The chance of identifying criticalities is higher.
Closed Card Sorting

Participants are provided with the content cards and the category cards and asked to place them in these given categories (which are pre-defined and cannot be modified). This method is usually used when adding new content to an existing site or reworking an open card sort’s insights.

Closed card sort helps gain better clarity of how the participants’ target market perceives these pre-defined headings. This ensures obtaining the confidence of adding content appropriately on a website or app.

This image has an empty alt attribute; its file name is Card-sorting_20201009_3-1024x677.png

Participants are not free to rename cards or make suggestions for new cards like open card sorting.

Closed card sorting helps with observational probing for the researcher as the pre-defined criteria for categorizing different items into groups have been set.

Remote Card Sorting

Here, an online software tool that provides the cards can work independently on sorting with their computers. This online software tool allows the user to set up and distribute it to as many participants as required.

Remote Card Sorting is unmoderated, which means there is no contact with the participant, so there is no way of understanding why the user has arranged the cards in a certain way.

These online software tools provide different ways to analyze the data. Many tools also enable recording the user’s screen and spoken feedback to facilitate the collation of qualitative data supporting quantitative metrics.

Reverse Card Sorting

Reverse card sorting (also called reverse lookup or tree testing) is a variation of closed card sorting where cards representing content, tasks, or navigation items are placed onto a predetermined hierarchy (or another type of structure). An assessment of how often users put the cards into the “right” categories on the hierarchy is done.

A participant is presented with a hierarchical diagram and a pile of cards representing information on categories and subcategories. The participant must place the cards at the correct level of the hierarchy. The cards are sorted into the right hierarchical location. This method helps validate changes to website navigation, hierarchical menus, or other task flows. It also helps compare the new navigation structure of a website to the old one.

This image has an empty alt attribute; its file name is Card-sorting_20201009_4-918x1024.png
Modified Delphi Method

This method relies upon the agreement of a group of participants in sequence to complete the card sorting exercise. The first participant provides their categorization based on the traditional open card sort methodology. Then the next participant can make any modifications to the first participant’s result or agree with it. The process is repeated in the loop with multiple participants until they reach a consensus on the result, and no further significant modifications are required.

The downside to this methodology is that any participant making significant changes to previous participants’ work can lead to an outlier, compromising the whole study’s results.

In-person Card Sorting

This method includes studies where an observer or researcher is physically present when the participants sort the cards. The researcher can ask further questions based on the participant’s choices and clarify any observations.

Participants are provided with a set of cards to move around and asked to talk through their thoughts and reasoning behind their decisions.

What are the generic steps in card sorting?

1. Choose the Medium and Topics

The first thing you need to determine in card sorting is how to perform it based on the type of card sorting chosen. Like on software or paper.

Each card should represent important content for the site; experts recommend 30–60 cards with no more than 70 topics.

Tip: Avoid the same words while naming topics because participants will tend to group those cards.

2. User organizes topics into groups

Shuffle the cards and give them to the participant. The participant must look at one card at a time from the pile (large or small). If the participant isn’t sure about a card, or its meaning, it can be left off. A set of “unknown” or “unsure” cards can exist rather than random cards.


  • The number of piles is not pre-set or predictable. The size depends on their mental models.
  • Users can change their minds as they work: they can move a card from one pile to another, split a pile into several new piles, merge two collections, etc. Card sorting is a bottom-up approach, and false starts are common.
3. User names the groups

After the participant has grouped all the cards, blank cards are handed over where he/she writes down a name for each group created. This step reveals the user’s mental model of the topic space. A few ideas for navigation categories may be sought, but it can’t be expected from participants to create useful labels.

Tip: Naming is essential after all the groups have been created so that the user doesn’t get locked into categories. The participant should be given the freedom to rearrange groups at any moment.

4. Debrief the user (optional but highly recommended)

Participants can be approached to explain the rationale behind the groups they’ve created. Helpful questions would include:

  • Which items were especially easy or difficult to place?
  • Did any items seem to belong in more than one group?
  • What thoughts do you have about any items left unsorted?

You can also ask the user to express their thoughts while performing the original sorting. Doing so provides insights, even though the analysis is time-consuming.

This information could push you into crosslinking from one category to the other or maybe also assigning the item to a definitive category if other reasons are leaning in that direction.

5. Get more Practical Group Sizes from the user 

The participant should not influence personal wishes or biases during the original sorting (steps 1–3). Once you define the user’s preferred grouping, and after the initial debrief, you can ask the participant to break large groups into smaller subgroups or vice versa.

6. Repeat with 15–20 users

The more the number of users, the easier it becomes to detect patterns in users’ mental models. We recommend a minimum of 15 participants for card sorting.

More number of users means getting diminishing returns for each additional user. The opposite means you won’t have enough data to reveal overlapping patterns in organization schemes.

7. Analyze the Data

After you have all the data, look for collective groups, category names or themes, and items frequently paired together. Other steps include:

  • Note the most grouped topics and unused cards
  • Understand the type of tags that are suggested by most of the participants
  • Figure out the logic participants use while they group the cards.
  • Review the participant comments, particularly regarding any other items of interest they may have brought up.
  • Combine patterns based on qualitative insights from a debriefing.

This will leave you better positioned to understand what the organization system will be most successful for your users.

Wrapping it up

Card sorting is highly recommended for a deeper understanding of the information architecture in UX design. It decodes the users’ understanding of your content.

Card sorting can be added with other information-architecture methods to identify issues in your category structure. Because at the end of the day, a well-structured Information Architecture promotes site/design/prototype popularity among end-users.

Are you interested in incorporating this technique for engaging UX design? Connect with our experts today.